In a previous post we had discussed how we can create mmwave & microwave frequency using a frequency comb.
https://akshatjiwannotes.blogspot.com/2024/11/making-broadband-generator-for.html
But there are a few problems with the method described in that post. Mostly related to fabrication of photodiodes on the substrate & its arrangement into a photonic pattern.
New method for generation of high frequency AC for communication and computing systems in gigabose to terrabose range.
Instead of using for whispering galleries or frequency combs fabricated on chips we use a diffraction grating.
The big advantage being :easy to fabricate ,easy to understand, simple operation, no complicated theory, works with uncollimated light source, stable reliable operation using passive components.
Photo detectors cavity resonators +diffraction grating :how it works
Light shines on a diffraction grating and forms a diffraction pattern
Photo diodes are placed on the pattern based on the frequency of signal to be generated
As light strikes the photodiode current is generated onto it and it travels on the transmission line.
The current from the transmission line is fed back into the light source to switch it on and off. This creates a feedback loop.
The spacing between patterns cause a delay in the arrival of light at photodiodes leading to a delay in current generation creating high frequency AC wave on the transmission line
Antennas can be used to filter out unwanted frequency auto generate signals before placing it on the transmission line
The only limitation in this setup is how small the photo diode can we make.
Some maths
d/t=s, speed
d/s=t,time
1/t=s/d=frequency (units bose)
Using these formulas above, simply by adjusting diffraction pattern spacing we can tune the frequency generated.
Suppose the distance d=3mm=3*10^-3m
Time t taken by light=3*10^-3/3*10^-8
=1*10^-11 sec
Frequency f=1/1*10^-11 bose=100gigabose
For d =3cm frequency f=10gigabose
The actual frequency generated will depend upon the switching speed of led.Most leds available today have a response time in 10s of nano seconds. Switching speeds in pico seconds will be needed to produce signals in gigabose range & even lower in terabose range. Such leds & associated circuits are not available today.
https://lednique.com/led-speed/
It's difficult to achieve picosecond or femtosecond fast switching speeds so we’ll have to come up with a different way to generate carrier frequencies.
Here’s what we'll do:instead of switching the light source we’ll delay the signal on the transmission line. The slight delay in arrival of signal will cause oscillations in the transmission line & create a standing wave of the desired frequency. This can then be filtered with a waveguide. Here are the steps:
Pass the light through a diffraction grating creating a diffraction pattern.
Now place 2 photodiodes on the diffraction pattern generated. The distance d represents the distance light has to travel to consecutive photodiodes. It will be chosen using the formula above to cause the required delay.
[The total delay will be sun of distance traveled by light and the distance traveled by the electrical current generated by photodiodes]
Next the two photodiodes are connected to the transmission line through a resistor. The resistance value of the o/p line is chosen such that it conducts only when the signal is present on both lines connecting the photodiodes.
PD1 receives the light signal earlier than PD2 which leads to a signal on line 1 before signal on line 2 and hence on the output transmission line there is an oscillating current that has a positive value only when both lines are active.
[When only signal 1 arrives there's no output on transmission line.]
This current will be oscillating at a frequency given by the inverse of time period at which light is incident on PD2 + the time it takes to travel from PD2 to the resistor through line 2.
Here we have combined the principles of both rc oscillator and delay line oscillator to create a light driven oscillating circuit.
[It is important to keep in mind here that electricity Travels at the speed of light.]
Some notes and questions
All you need to create an AC signal is to switch DC on and off at a particular frequency
Here we are exploiting the delay induced by light as it travels different distances. The signal generated would be one continuous fluctuating wave.
A Cavity magnetron is fundamentally an l c circuit. Like all LC circuits it converts DC to AC. The oscillating components are cavity resonators which are shaped to include inductive and capacitive elements. Electronic oscillation leads to the production of em wave.
Gunn diodes, impatt diodes generate wave forms using an LC circuit as an oscillator diodes are simply switching DC on and off.
Why are Quartz crystals used for clock generation at all if diodes can work with such a broad frequency spectrum?
What improvements does the photodiode setup offer over gunn diode / impatt diode material?architecture?
The first improvement is in materials. The same set of materials can be used both as light source & photodiodes due to the principle of optoelectronic reciprocity.
Second, we don't need to worry about creating complex geometry of cavities that can be as small as a few mm.
Lastly by adjusting the delay we can create different frequencies over a wide range. The flexibility is such that the medium will impose limitations on the propagation of waves rather than the device. Higher frequency waves will be more easily attenuated in the atmosphere.
.3-300 gigabose- this range sets limits on how far LC type oscillators can go. It seems good enough as it covers the entire spectrum of millimeter waves as well as centimeter waves, smaller than this and we are into terrabose.
What happens when a pulsed DC is applied to an LC oscillator?
How does that affect its time constant?
What happens when ac is applied to an LC oscillator?
What if the gunn / impatt diode was replaced by photodiodes in an lc circuit? How would it change anything?
Is the switching speed of gunn diode any higher than photodiode?
Can similar frequencies be expected from photodiode which similar response times if they are connected to similar circuits as in gunn and impatt diodes?
Can these wave forms we used as carrier waves?
Comments
Post a Comment